Factors Influencing the Non-recovery of Renal Function After the Relief of Urinary Tract Obstruction in Women with Cancer of Cervix

Abstract

Objectives. To identify factors associated with non-recovery of renal function after the relief of urinary tract obstruction in women with cancer of cervix. Patients and Methods. Thirty-seven women with cancer of cervix and obstructive renal failure were prospectively studied. Patients were evaluated in the day before nephrostomy (D0); and in the 1st (D1), 7th (D7), and 30th day (D30) after unilateral nephrostomy. The following parameters were analyzed: age serum urea, and serum creatinine (S_creat), in D0, D7, and D30. Renal cortical thickness was measured by ultrasonography in D0. During nephrostomy procedure, urine samples for urinalysis and culture were collected and renal biopsy was attempted. Patients were divided in two groups according to S_creat in D30: patients with S_creat <1, 4 mg/dL (R group) and those with S_creat ≥ 1, 4 mg/dL (nR group). Results. Ten patients died before D30 and were not studied. R group (n = 12) was younger (43 ± 9 vs. 52 ± 10 years, p<0.05) than nR group (n = 15), and presented lower serum urea in D0 (134 ± 67 vs. 212 ± 51 mg/dL, p<0.01) and in D7 (94 ± 20 vs. 155 ± 18 mg/dL, p<0.05). Renal cortex was thicker in R group (17 ± 1 vs. 13 ± 1 mm, p<0.05). Survival was higher in R group (11/12) than in nR group (0/15, p<0.01). All urine samples showed leukocyturia but culture was positive in only two. In three out of nine biopsy attempts enough material was obtained. Light microscopy revealed interstitial fibrosis plus tubular atrophy, and partial glomerular fibrosis was present in two biopsies. Conclusions. The non-recovery of renal function after the relief of obstruction in human obstructive renal failure is associated with older age and decreased renal cortical thickness. The complete renal recovery improves patient survival despite malignancy.

• Obstructive renal failure
• Urinary tract obstruction
• Cancer
• Renal functional recovery
• Humans
• Renal biopsy

Introduction

Urinary tract obstruction is a common cause of renal failure. Complete renal recovery can be achieved after an early relief of the obstruction.[[1]] Urinary tract obstruction can occur in any age but it is more common in childhood and in old age. Malignancy is a common cause of urinary tract obstruction among adults, either males or females.[[2]] The most important factors determining functional recovery are severity and duration of obstruction.[[3]], [[4]] Experimental studies have shown that functional recovery depends on animal species, obstruction duration and, in cases of unilateral obstruction, on the presence of a normal contralateral kidney.[[5]] In man, partial functional recovery was reported after complete unilateral ureteral obstruction for three months and even after more than seven months of complete bilateral obstruction.[[6]], [[7]] However, it is very difficult to precise in humans the obstruction duration and for how long the obstruction is complete. Therefore, it is important to look for factors associated with functional non-recovery in order to avoid invasive procedures such as nephrostomy or “JJ” ureteric stent, on kidneys with low potential for recovery. Some parameters such as renal cortical thickness, degree of hydronephrosis, presence of anuria—that could be an indication of complete urinary obstruction—and intensity of interstitial fibrosis, could be markers of renal functional non-recovery.[[8]], [[9]], [[10]], [[11]] In order to identify the association of these factors with renal functional non-recovery, women with advanced cancer of cervix, and obstructive renal failure were studied after unilateral nephrostomy.

Patients and Methods

Thirty-seven successive women with cancer of cervix stage III or IV and obstructive renal failure were prospectively studied while hospitalized in Hospital Pérola Byngton, São Paulo, SP, Brazil, for unilateral nephrostomy tube placement. Nephrostomy was the procedure adopted to relieve the obstruction as “JJ” stent was not available in that hospital. All patients with clinical conditions to be submitted to the surgical procedure were eligible for nephrostomy tube placement. Diabetic or hypertensive patients were not studied because of possible previous renal impairment. Patients were studied in the day before nephrostomy (D0), and in 1st (D1), 7th (D7), and 30th day (D30) after nephrostomy. The following parameters were recorded: age, need for dialysis before nephrostomy, and prior treatment of the cancer with radiotherapy. In D0, D7, and D30, serum urea (S_urea) and serum creatinine (S_creat) were measured. Because some patients needed dialysis before nephrostomy, the peak value of S_urea and S_creat before D0 was also recorded. Urinary volume was measured in D0 through bladder catheterization and in D1 and in D7 through the nephrostomy tube. Anuria was defined as urinary volume <50 mL/24 h. In D0 and D7 the degree of hydronephrosis (mild, moderate, or severe) and renal cortical thickness were determined by ultrasonography. A nephrostomy tube was placed percutaneously by two experienced urologists with local anesthesia and under ultrasonographic guidance. The kidney to be drained was chosen by the urologists on their own preferences. Urine samples were collected from the drained kidney just after nephrostomy tube placement, for urinalysis in 13 patients and for culture in 12 patients. Renal biopsy was attempted by one or two punctures during the nephrostomy procedure in nine patients. Renal specimens were dyed with hematoxylin–eosin and Masson and analyzed by light microscopy. Patients were followed at least for 35 days or until death. Patients were divided in two groups according to S_creat value in D30: patients with normal S_creat (<1.4 mg/dL, R group) and patients with S_creat ≥ 1.4 mg/dL (nR group). This study was approved by the Ethics Committee of Hospital Pérola Byngton and Hospital das Clínicas. All patients gave their written informed consent. Statistical analysis was performed using GraphPad Prism software version 2.01, 1996. Mann–Whitney test was used for the analysis of quantitative variables, and Fischer's test for the qualitative variables. Survival curve was calculated by logrank test. Statistical significance was considered for p<0.05. Data are presented as mean ± SD or expressed as percentage.

Results

Ten patients died before D30 and their data were not included. Nevertheless it is noteworthy that their data until D7 were not different when compared with patients who survived longer. Among the 27 patients alive in D30, 15 patients did not present normal S_creat (nR group). The remaining 12 patients (44%) constituted the R group. R group was younger than nR group: 43 ± 9 vs. 52 ± 10 years, p = 0.025. Prevalence of prior radiotherapy was similar in both groups: 58% in R group and 53% in nR group, p = 0.55. Urinary volume was similar in both groups in D1 (1,889 ± 352 mL/24 h in nR group and 2,650 ± 673 mL/24 h in R group, p = 0.467) but in D7 it was higher in R group: 2,986 ± 497 mL/24 h vs. 1,607 ± 285 mL/24 h, p = 0.04. Prevalence of anuria in D0 was similar in both groups: 17% in R group and 20% in nR group, p = 0.44. Ultrasonography showed that cortex of the drained kidney was thicker in R group: 16.96 ± 1.05 mm vs. 13.22 ± 0.78 mm, p<0.01. All patients with cortical thickness of the drained kidney ≤12 mm were in nR group. However, the degree of hydronephrosis in the kidney subjected to nephrostomy was not different in D0 between the two groups: severe in 21% of nR group vs. 17% of R group, p = 0.39. In D7, ultrasonography was performed in eight patients of R group and in 11 patients of nR group. It was found that even in the presence of a well placed and working nephrostomy in all patients, only one patient of R group and four of nR group showed hydronephrosis reduction in the homolateral kidney (p = 0.34). In the contralateral kidney, the degree of hydronephrosis decreased in five patients of R group and in three patients of nR group (p = 0.18).Although the prevalence of need for dialysis before nephrostomy was similar in both groups (40% in nR group and 42% in R group, p = 0.62), S_urea in D0 was lower in R group as shown in Table 1. S_urea was also lower in R group in D7 and in D30. However, lower S_creat values in R group were found only in D30. Urinalysis showed leukocyturia in all samples but culture was positive in only two. Urinary pH was similar in both groups: 5.7 ± 0.6 in R group and 6.0 ± 1.0 in nR group (p = 0.83), likewise, specific gravity: 1012 ± 3 in R group and 1013 ± 5 in nR group (p = 0.67). During the follow-up nine patients in nR group died compared with only one in R group (p<0.01) and patient survival was longer (p<0.01) in R group as shown in Fig. 1. Death was always caused by the spread of the malignancy. Attempts to obtain renal tissue for histological analysis were done in nine successive patients but enough material was obtained in only three. All samples presented no more than three glomeruli. Because of the difficulties to obtain adequate histological material, it was decided not to further perform renal biopsies. All three biopsies showed diffuse and severe tubular atrophy plus interstitial fibrosis. In two biopsies glomeruli showed partial fibrosis. Acute tubular necrosis was not found. Although important renal histological scarring was present, two of these patients normalized S_creat (R group) and the other died in the 13th day after nephrostomy with S_creat = 7.5 mg/dL. In 10 patients, it was observed normal S_creat prior to nephrostomy despite the presence of ultrasonographic signs of bilateral urinary tract obstruction. Three of these patients died before D30, for the other seven patients the period of time between the recorded normal value of S_creat and the nephrostomy insertion was similar in both groups: 25 ± 3 days in four patients of R group and 32 ± 12 days in three patients of nR group (p = 0.63).

Table 1. Evolution of serum urea and serum creatinine values in R and nR groups

	Serum urea (mg/dL)		Serum creatinine (mg/dL)
	R group	nR group	R group	nR group
Peak	223 ± 16	263 ± 23	13.1 ± 2.6	10.4 ± 1.7
	n = 11	n = 15	n = 11	n = 15
D0	134 ± 67	212 ± 51a	8.0 ± 4.3	8.3 ± 3.9
	n = 11	n = 14	n = 12	n = 15
D7	94 ± 20	155 ± 18b	4.0 ± 4.2	4.8 ± 3.4
	n = 12	n = 14	n = 12	n = 15
D30	48 ± 25	108 ± 60a	0.9 ± 0.2	3.6 ± 3.2
	n = 11	n = 15	n = 12	n = 15

D0, D7, and D30: respectively, the day before nephrostomy, 7th and 30th days after the nephrostomy.

^aP<0.01 vs. R group.

^bP<0.05 vs. R group.

Figure 1. Survival curves in R and in nR group. R group is represented as squares and straight lines and nR group as triangles and interrupted lines.

Discussion

Duration of urinary obstruction is crucial to achieve renal function recovery after the relief of the obstruction in experimental models.[[5]] In dogs, non-recovery was observed after relief of unilateral obstruction of five or six weeks' duration.[[4]], [[12]] In humans, little or no improvement should be expected after more than 12 weeks of complete obstruction.[[3]] However, it was reported partial functional recovery—enough for stopping dialysis—even after obstruction of seven months' duration.[[7]] In humans it is very difficult, or even impossible, to precise the duration of a functionally important urinary obstruction. It is also very important to identify other factors that can predict non-recovery after relief of obstruction in order to avoid procedures, e.g., nephrostomy or “JJ” ureteric stent, in kidneys with low potential for recovery, that may only lead to decreased quality of life.[[13]], [[14]], [[15]]

Olxd age has been reported as predictor of poor functional recovery after acute renal failure due to acute tubular necrosis.[[16]] Studies of renal function after relief of urinary obstruction caused by enlarged prostate have shown that older patients had poorer and later functional recovery.[[11]], [[17]] In the present study, age was an important determining factor for non-recovery: nR group was older than nR group. The influence of age in recovery is probably due to the loss of renal functional reserve as consequence of aging.

It was reported that the presence of urinary pH ≤ 6.0 is associated with complete radiological recovery after relief of unilateral obstruction.[[18]], [[19]] In the present study no differences in urinary pH were found between R and nR groups. However, the present patients were different from those previously reported because they had severe renal failure and bilateral urinary obstruction. Thus, in patients with these conditions—bilateral obstruction plus renal failure—urinary pH is not a good predictor of functional recovery. The presence of infection was also reported to worsen functional recovery.[[11]] No conclusion about the influence of urinary infection on functional recovery could be drawn in the present study because of the low rate of urinary infection: only two positive cultures were obtained, one from each group. However, all urine samples presented leukocyturia. A possible source for the increased number of white cells found in urine may be interstitial infiltrating leukocytes, whose presence is well described in experimental obstruction.[[20]]

Nadig and Valk studying nine patients after the relief of obstruction caused by prostate enlargement reported that the later stable glomerular filtration rate depended on the last glomerular filtration value found prior the relief of obstruction.[[21]] However, this finding was not verified in the present study: as presented in Table 1, the peak S_urea and S_creat values were similar in both groups as well as the prevalence of need for dialysis before nephrostomy. In these patients, glomerular filtration rate is so reduced that S_urea or S_creat measurements do not detect minor differences. Alternatively, prior glomerular filtration loss may not influence renal recovery after relief of obstruction. More studies are needed to clarify this point.

In patients with renal failure, ultrasonography is the method of choice for the diagnosis of urinary obstruction.[[22]], [[23]] Some ultrasonographic findings were pointed as predictors of poor or no functional recovery: presence of small kidneys, increased parenchymal echogenicity, and thin cortex, but no cutoff for these parameters was determined.[[11]] Objective measurements of echogenicity based on whole kidney dimensions are difficult to perform. Also, in the presence of severe hydronephrosis the kidney's major diameter can be normal or even increased. Given the above considerations, we opted to measure cortical thickness. A clear difference in this parameter was found between the two groups, nR group showed thinner cortex. The presence of normal cortical thickness may not be a predictor of functional recovery as reported by Coroneos et al.[[9]] However, as demonstrated in the present study, the presence of a cortical thickness ≤12 mm is probably a good indicator of poor or no functional recovery. Larger studies are necessary for the establishment of the width of cortical thickness that securely rules out the functional recovery of an obstructed kidney. The renovascular resistive index measured by Doppler ultrasonography has shown low specificity and sensitivity for detecting obstruction, but it could otherwise be used as a recovery predictor in clearly obstructed kidneys.[[24]] Among the imaging parameters, DMSA scintigraphy has a good accuracy in predicting recovery after relief of obstruction.[[25]] Nevertheless, especially for low-income patients, such expensive procedure can be avoided if prediction of functional recovery by imaging could be appropriately accomplished by ultrasonography.

The gold standard method for prediction of recovery is the histological renal examination as Coroneos et al. demonstrated in two cases.[[9]] However, in the present study it was extremely difficult to obtain appropriate material for analysis, forcing us to abandon renal biopsy attempts. As pointed out, in only three out of nine attempts, enough material was obtained and with no more than three glomeruli in each biopsy. Histological analysis of this material showed that even with important renal scarring, two patients achieved normal S_creat. Acute tubular necrosis, that could affect prognosis because of its potential reversibility, was not found in any biopsy. Thus, renal biopsy is a risk procedure that is not recommended in presence of hydronephrosis.

Another finding deserves to be remarked: in the 10 patients for whom the time between normal S_creat and severe renal failure requiring diversion could be determined, the time elapsed was very short, about one month. Thus, it is important that upon detection of urinary obstruction, even with normal S_creat, early intervention should be carried out to avoid rapid occurrence of renal failure.

Experimentally it has been demonstrated that the presence of a working kidney can suppress the recovery of the injured one: the so termed “renal counterbalance.”[[5]] If this mechanism applies to man, the contralateral kidney should diminish its glomerular filtration and possibly decrease its degree of hydronephrosis, after successful diversion as occurred in R group. This in fact may have happened: despite no statistical significance (p = 0.18), the ultrasonography performed in D7 showed that the degree of contralateral hydronephrosis decreased in five of the eight evaluated patients of R group but in only three of the 11 evaluated patients of nR group.

In patients with malignancies, normalization of S_creat after unilateral relief of urinary obstruction occurs in 42%, very similar to the present results: 44%.[[26]] Despite high mortality in the first month after nephrostomy (27%), an improved survival was observed in R group. Survival duration after urinary diversion in urinary obstruction secondary to malignancy, has been associated with type of malignancy,[[27]], [[28]] spread of the disease[[29]] and possibility of further treatment.[[13]], [[28]], [[30]] In patients with prostate cancer and urinary obstruction, better survival was observed in those patients for whom further cancer therapy was still available after the urinary decompression.[[30]] As Fallon et al.[[13]] pointed out, the possibility of further therapy is an important factor determining survival. However, in the present study 58% of R group patients and 53% of nR group had already undertaken radiotherapy. Furthermore, all patients with no prior radiotherapy undertook such treatment after nephrostomy. This observation indicates that the improvement in R group survival cannot be ascribed only to further available treatment.

In conclusion, the present study has demonstrated that in bilateral urinary obstruction secondary to cancer of cervix: 1. Functional non-recovery after the relief of obstruction can be expected for older age and for patients with reduced cortical thickness; 2. Complete renal functional recovery improves survival despite unavailability of further cancer therapy.

Acknowledgment

The authors would like to thank Drs. Miguel Conrado Leal and Riberto Liguori, urologists of Hospital Pérola Byngton, that performed all nephrostomies and to Dr. Denise M. A. C. Malheiros, pathologist of Hospital das Clínicas, for the histological analysis.